Distributed-Type Markov Chain Monte Carlo

1. A method, comprising: generating, via a swap proposal core, a plurality of swap proposals associated with a distributed-type Markov chain Monte Carlo procedure; receiving, via a first chain core, one or more of the swap proposals from the swap proposal core; generating, via the first chain core, a plurality of random numbers; processing, via the first chain core, a current iteration of a first chain portion of the distributed-type Markov chain Monte Carlo procedure based at least in part on the plurality of random numbers; waiting, via the first chain core, for a current iteration of a second chain portion of the distributed-type Markov chain Monte Carlo procedure associated with a second chain core to reach a corresponding iteration based at least in part on one or more of the swap proposals; and wherein the generation of random numbers, via the first chain core, proceeds without interruption during the waiting, via the first chain core, for the second chain core to reach the corresponding iteration of the distributed-type Markov chain Monte Carlo procedure.

2. The method of claim 1 , wherein the generation of the plurality of swap proposals, via the swap proposal core, proceeds without interruption during the waiting, via the first chain core, for the second chain core to reach the corresponding iteration of the distributed-type Markov chain Monte Carlo procedure.

3. The method of claim 1 further comprising: storing, via a swap proposal generation queue, one or more of the swap proposals; and distributing, via the swap proposal core, one or more of the swap proposals to the first chain core from the swap proposal generation queue.

4. The method of claim 1 further comprising storing, via a swap proposal distribution queue, one or more of the swap proposals received via the first chain core, wherein the swap proposal distribution queue is associated with the first chain core and a second swap proposal distribution queue is associated with the second chain core.

5. The method of claim 1 , further comprising storing, via a random bit queue, one or more of the plurality of random numbers, wherein the random bit queue is associated with the first chain core and a second random bit queue is associated with the second chain core.

6. The method of claim 1 , after the waiting, via the first chain core, for the second chain core to reach the corresponding iteration of the distributed-type Markov chain Monte Carlo procedure, the method further comprising: distributing, via the swap proposal core, further swap proposals to the first chain core and the second chain core; determining, via the first chain core and/or the second chain core, to swap a state associated with the first chain portion with a state associated with a second chain portion based at least in part on one or more of the swap proposals stored via the swap proposal distribution queue; and processing, via the first chain core, a subsequent iteration of the first chain portion based at least in part on the state swapped from the second chain as well as the plurality of random numbers stored via the random bit queue.

7. The method of claim 1 , wherein the distributed-type Markov chain Monte Carlo procedure comprises a Metropolis-coupled-type Markov chain Monte Carlo procedure.

8. The method of claim 1 , wherein a specific swap proposal comprises information related to the first chain portion, the second chain portion, and a specific iteration.

9. The method of claim 1 , wherein the method is adapted to preserve the statistical performance of the distributed-type Markov chain Monte Carlo procedure.

10. A multicore processor including a plurality of processor cores, the multicore processor comprising: a swap proposal core configured to generate a plurality of swap proposals associated with a distributed-type Markov chain Monte Carlo procedure; a first chain core configured to receive one or more of the swap proposals from the swap proposal core, generate a plurality of random numbers, process a current iteration of a first chain portion of the distributed-type Markov chain Monte Carlo procedure based at least in part on the plurality of random numbers, and wait for a current iteration of a second chain portion of the distributed-type Markov chain Monte Carlo procedure associated with a second chain core to reach a corresponding iteration based at least in part on one or more of the swap proposals; and the second chain core configured to process a current iteration of the second chain portion of the distributed-type Markov chain Monte Carlo procedure, wherein the first chain core is further configured to continue generation of random numbers without interruption during the wait for the second chain core to reach the corresponding iteration of the distributed-type Markov chain Monte Carlo procedure.

11. The multicore processor of claim 10 , wherein the swap proposal core is further configured to continue generation of the plurality of swap proposals without interruption during the wait for the second chain core to reach the corresponding iteration of the distributed-type Markov chain Monte Carlo procedure.

12. The multicore processor of claim 10 , further comprising: a swap proposal generation queue configured to store one or more of the swap proposals, and wherein the swap proposal core is further configured to distribute one or more of the swap proposals to the first chain core from the swap proposal generation queue.

13. The multicore processor of claim 10 , further comprising: a swap proposal distribution queue configured to store one or more of the swap proposals received via the first chain core, wherein the swap proposal distribution queue is associated with the first chain core; and a second swap proposal distribution queue configured to store one or more of the swap proposals received via the second chain core, wherein the second swap proposal distribution queue is associated with the second chain core.

14. The multicore processor of claim 10 , further comprising: a random bit queue configured to store one or more of the plurality of random numbers, wherein the random bit queue is associated with the first chain core; and a second random bit queue configured to store one or more of the plurality of random numbers, wherein the second random bit queue is associated with the second chain core.

15. A computing device including a plurality of processor cores, the computing device comprising: a multicore processor, comprising: a swap proposal core configured to generate a plurality of swap proposals associated with a distributed-type Markov chain Monte Carlo procedure, a first chain core configured to receive one or more of the swap proposals from the swap proposal core, generate a plurality of random numbers, process a current iteration of a first chain portion of the distributed-type Markov chain Monte Carlo procedure based at least in part on the plurality of random numbers, and wait for a current iteration of a second chain portion of the distributed-type Markov chain Monte Carlo procedure associated with a second chain core to reach a corresponding iteration based at least in part on one or more of the swap proposals, the second chain core configured to process a current iteration of the second chain portion of the distributed-type Markov chain Monte Carlo procedure, and wherein the first chain core is further configured to continue generation of random numbers without interruption during the wait for the second chain core to reach the corresponding iteration of the distributed-type Markov chain Monte Carlo procedure; a memory bus; and a system memory configured in communication with the multicore processor via the memory bus.

16. The computing device of claim 15 , wherein the swap proposal core is further configured to continue generation of the plurality of swap proposals without interruption during the wait for the second chain core to reach the corresponding iteration of the distributed-type Markov chain Monte Carlo procedure.

17. The computing device of claim 15 , wherein the swap proposal core further comprises a swap proposal generation queue configured to store one or more of the swap proposals, and wherein the swap proposal core is further configured to distribute one or more of the swap proposals to the first chain core from the swap proposal generation queue.

18. The computing device of claim 15 , wherein the multicore processor further comprises: wherein the first chain core further comprises a swap proposal distribution queue configured to store one or more of the swap proposals received via the first chain core; wherein the second chain core further comprises a second swap proposal distribution queue configured to store one or more of the swap proposals received via the second chain core; wherein the first chain core further comprises a random bit queue configured to store one or more of the plurality of random numbers, wherein the random bit queue is associated with the first chain core; and wherein the second chain core further comprises a second random bit queue configured to store one or more of the plurality of random numbers.

19. An article comprising: a non-transitory medium comprising machine-readable instructions stored thereon, which, if executed by one or more processors, operatively enable a computing device to: generate, via a swap proposal core, a plurality of swap proposals associated with a distributed-type Markov chain Monte Carlo procedure; receive, via a first chain core, one or more of the swap proposals from the swap proposal core; generating, via the first chain core, a plurality of random numbers; process, via the first chain core, a current iteration of a first chain portion of the distributed-type Markov chain Monte Carlo procedure based at least in part on the plurality of random numbers; wait, via the first chain core, for a current iteration of a second chain portion of the distributed-type Markov chain Monte Carlo procedure associated with a second chain core to reach a corresponding iteration based at least in part on one or more of the swap proposals; and wherein the generation of random numbers, via the first chain core, proceeds without interruption during the waiting, via the first chain core, for the second chain core to reach the corresponding iteration of the distributed-type Markov chain Monte Carlo procedure, and wherein the generation of the plurality of swap proposals, via the swap proposal core, proceeds without interruption during the waiting, via the first chain core, for the second chain core to reach the corresponding iteration of the distributed-type Markov chain Monte Carlo procedure.

20. The article of claim 19 , further operatively enabling the computing device to: store, via a swap proposal generation queue, one or more of the swap proposals; and distribute, via the swap proposal core, one or more of the swap proposals to the first chain core from the swap proposal generation queue; store, via a swap proposal distribution queue, one or more of the swap proposals received via the first chain core, wherein the swap proposal distribution queue is associated with the first chain core and a second swap proposal distribution queue is associated with the second chain core; and store, via a random bit queue, one or more of the plurality of random numbers, wherein the random bit queue is associated with the first chain core and a second random bit queue is associated with the second chain core.